open-llm-leaderboard/details_TheBloke__Guanaco-3B-Uncensored-v2-GPTQ
收藏数据集概述
数据集简介
该数据集是在对模型 TheBloke/Guanaco-3B-Uncensored-v2-GPTQ 进行评估时自动创建的,用于 Open LLM Leaderboard。
数据集结构
- 数据集包含 64 个配置,每个配置对应一个评估任务。
- 数据集从 2 次运行中创建,每次运行在每个配置中作为一个特定的分片,分片名称使用运行的时间戳。
- "train" 分片始终指向最新的结果。
- 一个额外的配置 "results" 存储所有运行的聚合结果,用于计算和显示 Open LLM Leaderboard 上的聚合指标。
数据加载示例
python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_TheBloke__Guanaco-3B-Uncensored-v2-GPTQ", "harness_winogrande_5", split="train")
最新结果
以下是 2023-10-29T01:04:16.242483 运行的最新结果: python { "all": { "em": 0.0045092281879194635, "em_stderr": 0.0006861346899095007, "f1": 0.06708368288590627, "f1_stderr": 0.0016014292768729186, "acc": 0.322384038037953, "acc_stderr": 0.0072675866532889944 }, "harness|drop|3": { "em": 0.0045092281879194635, "em_stderr": 0.0006861346899095007, "f1": 0.06708368288590627, "f1_stderr": 0.0016014292768729186 }, "harness|gsm8k|5": { "acc": 0.001516300227445034, "acc_stderr": 0.0010717793485492612 }, "harness|winogrande|5": { "acc": 0.6432517758484609, "acc_stderr": 0.013463393958028728 } }
配置详情
以下是数据集的配置详情:
-
harness_arc_challenge_25
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|arc:challenge|25_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|arc:challenge|25_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_drop_3
- 分片:2023_10_29T01_04_16.242483
- 路径:
**/details_harness|drop|3_2023-10-29T01-04-16.242483.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|drop|3_2023-10-29T01-04-16.242483.parquet
- 路径:
- 分片:2023_10_29T01_04_16.242483
-
harness_gsm8k_5
- 分片:2023_10_29T01_04_16.242483
- 路径:
**/details_harness|gsm8k|5_2023-10-29T01-04-16.242483.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|gsm8k|5_2023-10-29T01-04-16.242483.parquet
- 路径:
- 分片:2023_10_29T01_04_16.242483
-
harness_hellaswag_10
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hellaswag|10_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hellaswag|10_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_5
- 分片:2023_10_03T21_39_11.409465
- 路径:多个文件路径
- 分片:latest
- 路径:多个文件路径
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_abstract_algebra_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_anatomy_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_astronomy_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_business_ethics_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_clinical_knowledge_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_college_biology_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_college_chemistry_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_college_computer_science_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_college_mathematics_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_college_medicine_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_college_physics_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_computer_security_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_conceptual_physics_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_econometrics_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:2023_10_03T21_39_11.409465
-
harness_hendrycksTest_electrical_engineering_5
- 分片:2023_10_03T21_39_11.409465
- 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-10-03T21-39-11.409465.parquet
- 路径:
- 分片:latest
- 路径:`**/details_harness|hendrycksTest-electrical_engineering|5_2023-10-
- 分片:2023_10_03T21_39_11.409465



